UW-16(f)(3)(a)

[-CMF]

I am needing help with the use of O'lets on Sec. VIII Div. 1 vessels when the shell thickness is greater than .375". My thoughts are they aren't allowed per UW-16(f)(3)(a)when welded from the outside only. Many people argue its full pen. buts its full pen thru the fitting followed up with a fillet weld in my eyes. My reasoning is that the fitting isn't of uniform thickness thus being connected per Fig. UW-16.2 sketch (m). I am rejecting some venders drawings and calcs. for this and getting a lot of flak. I have spoken to my AIS buddy and he has agreed with me. Compress agrees as well if you check the coupling box instead of the uniform thickness option when designing a new nozzle. I have my venders change to a threaded bosset fitting and I feel this is allowed when welded from the outside only on shells thicker than .375" since the fitting is of uniform diameter and thickness so when designing it in Compress you can check the uniform thickness box and install as a set on. It will default to Fig. UW-16.1 sketch (a) and I am OK with that. Your thoughts?

 

[XL83NL]

I believe an olet doesnt fall under UW-16 as it's not listed in table U-3. Second, why would you consider to bring an olet under UW-16(f) anyway? Olets require full pen welds per their design; UW-16(f)(3)(-a) is for fillet welded attachments (or im missing something here). Id use UW-16(f) to weld a B16.11 half coupling with a fillet only to a vessel's shell.
An olet should always be welded full pen., they're designed as such. But they're actually designed for piping systems. Not that you cant design them for a vessel, but I dont see why you'd wanna use them on a vessel. Cant see why an alternative would be less economical.

 

[-CMF]

thanks for your response. I consider UW-16(f) since its welded on the outside of the shell only and is finished by a fillet weld. Yes, its designed for full pen, but the full pen is thru the fitting not the shell. A TOL actually falls under MSS SP-97. I don't want to use them (TOL's), my venders do and I am saying use the coupling or a bosset set-on. Their design (TOL's) is such that they are for piping as Boney Forge and others will clarify in their integral reinforcement statements that they comply with B31.1 and B31.3. Sec. VIII isn't listed. They (Boney Forge) will run calcs. for Sec. VIII but they are just for area replacement which we know is exempt per UG-36(c)(3) as well as weld strength per UW-15(b)(2). This is why I believe TOL's fall under UW-16(f)(3)(a) when attached on shells thicker than .375" and welded from the outside only. Please let me know your thoughts.

 

[XL83NL]

Im not following you. Any 'OL' (weldolet, flexolet, pipet, elbolet, flangolet) is, as far as I know, designed to MSS SP-97, which isnt a listed standard under VIII-1. UW-16(f)(3) is for attachment by fillet weld, what youre doing is something else. Im not understanding what youre trying to justofy. Why use an olet fitting which has a huge amount of weld metal (and thus potential problems for warping your vessel), when you can do this much more easy with a piece of pipe, additionally extended with a repad, or just a simple half coupling?

 

[-CMF]

I agree with you and in essence you are answering my questions. Yes, they aren't listed in table U-3 but my supplier has designed the vessel on a 2" shell to have one. Within Compress, you have three choices when designing a nozzle; uniform thickness (which an O'let is not),variable thickness (which is HB type nozzles, etc.) or couplings. If you select uniform thickness as a set on Compress will default to Fig. UW-16.1 sketch (a) but its not uniform thickness so you must select the coupling option. When doing so, it defaults to a set-on and once you run the calcs, its kicked per UW-16(f)(3)(a). My thoughts are a O'let is per Fig. UW-16.2 sketch (m) and if I am right, the bottom of the page says see the limitations of UW-16(f). An O'let is finished with a fillet and another key factor is the wording about being welding on the outside of the shell only. Its full pen., but not thru the shell. Your thoughts?

 

[marty007]

Just because Compress doesn't have functionality to properly model an attachment detail doesn't mean it can't be used.

In essence, an olet fitting is a cylindrical component often made from forged material. Per code, as long as it is supplied with an acceptable mill-cert, and you can calculate the reinforcement, there is nothing wrong with using them on a pressure vessel. The attachment is made per Fig UW-16.1(a), and the calculation can be made per UG-37 (likely exempt per UG-36(c)(3)(a)).

If you do need to perform calcs per UG-37 (large olet), then it's something you'll have to do outside of Compress because of their geometry.

 

[-CMF]

thank you marty007. I agree with your comments pertaining to Compress but feel it does have functions to fully calculate an O'let if the user enters the correct information. I would disagree with you that an O'let is not per Fig. UW-16.1 sketch (a) but in fact per Fig. UW-16.2 sketch (m). This is why I am rejecting its connection on my above mentioned vessel. My O'let is under 3" NPS (3/4" NPS actually). Another point I feel would differ from yours is that an O'let is not a cylindrical component since if so, it would be of uniform diameter and thickness whereas an O'let is not.

 

[cbPVme]

UW-16.2(m) is not a full penetration weld; it's a partial groove with filet. Olets should be attached with a full penetration weld per UW-16.1(a). Typically an olet comes with an angled bevel. The area under this is filled in and would be considered full penetration. If you have a copy of ASME B31.3, Figure 328.5.4F has a really good detail on how an o-let should be welded.

Since they're not uniform thickness, if you have a situation where UG-37 calculations are required, I think you would have to physically measure the cross sectional area of the olet.